Subject and object are functional roles. There is no guarantee that either has any stronger unity than is required by its role. Referentially, members of a subject-object pair may pick out parts of the same content. This can result in confusion when terms are used at different levels of analysis.
In the Sociology of Knowledge? post, I complained about a naive, unproblematic distinction between mind and world, then went on to speak of an asymmetric mutual determination. The latter sort of language might standardly be taken to imply a relation between distinct things, contradicting the former language. However, in context, the latter phrase is intended to be anaphoric at a higher level. In this case, mutual determination and the lack of an unproblematic distinction are two ways of talking about the same state of affairs.
This sort of mixed-metaphor-like phenomenon leading to apparent literal inconsistency often crops up when different dialectical levels are mentioned. We have to choose between potentially cumbersome formal disambiguation and extra interpretive work. (See also Aristotelian Dialectic.)